During the heady, pre-IPO days of Facebook, the tech world was being told that because half a billion people were registered on the site, not only would it become the platform for virtually any business venture or marketplace, it would also harness the contents of its massive data farms to drive customers to the right advertisers. The concept was that mapping out the connections and likes between people and their favorite pages could target ads to precisely fit someone’s profile and served as part of a news feed or a handy sponsored link in the sidebar. This social web was supposed to be the way, the word, and the light of all social media to a viable business model. If we go by the performance of Facebook’s stock, that’s not exactly what happened, and the rather mediocre returns on investment advertisers are seeing on the site only add insult to injury. Just as many users click on blank ads as they do real ones, suggesting that when they do click on a sponsored link, it’s likely to be a mistake. So much for hypertargeting as the future of social.
Here’s the problem a lot of large social media sites trying to make money can’t seem to solve. All the millions of users they have are so used to having everything free and have gotten so adept at filtering ads screaming at them from every direction, they’re using the tool for its social benefit while treating what’s now supposed to be the site’s bread and butter as an annoyance they just have to tolerate to use the awesome free tool. This is why Facebook or Twitter can’t even joke about a paid subscription to keep using their sites. Maybe about 1% of hardcore users who do need to keep using their tools will switch over while the other 99% flock elsewhere. If people are not used to paying for something, they sure aren’t going to start unless there’s simply no way to get what they want otherwise, and even then, many will just choose to go without, figuring that a monthly or even annual fee is asking too much for how little time they spend on these sites. And this is why we were being pitched the social web as the answer to the question of how any social media site should make enough money to truly live up to its hype. Instead of selling a small and controversial monthly subscription, they’ll just sell your data to companies instead.
However, the social web concept also suffers from an underlying flaw. It doesn’t account for a lot of human behavior because the interactions on the site effectively obscure them. You might get an ad for Nike running shoes because you liked a comment by your friend about how much she loves running in her new Nikes. From a computer standpoint, it makes perfect sense. You were friends for years, you comment on each other’s’ posts on a regular basis, you must be at least a little swayed by your friend’s opinion, right? But how does Facebook know that you’re friends on the site only because your parents know her parents and it would look bad if you declined her friend request? Where do you specify that you are on a completely different part of the political spectrum than her, couldn’t care less about running or tennis shoes, and only interact with her for the sake of appearances? She could say that Nike shoes cured her guinea pig’s cancer and you still wouldn’t care, mostly because you just ignored her status update in your feed, like you do for nearly every one of her posts. In this case, the social graph delivers a dud.
Human behavior is a very complex subject and there’s a wide variety of reactions humans have to the same ideas, events, and things. Trying to reduce them into predictable formulas to build a graph designed to predict how they’ll act is a fool’s errand. It will either be too specific, or vague to the point of being utterly useless. You’ll be relying on the belief that people are exactly who they are and what they are online, a real stretch considering that we have trouble figuring out if people are who they say they are or believe what they say they do in person. Add to that all the changes we experience throughout our lives, changes shaped by hormones, big decisions and their consequences, and even small events that inspire us to do something new and different at a moment’s notice, and you may as well try to call your social web Laplace since it would have to summon a variation of its metaphorical demon to get any sort of consistent accuracy when put into action. To ignore this was either wishful thinking on Facebook’s part, or an complete lack of familiarity with how humans generally tend to work out in the wild of the internet and IRL…
[ illustration by Bogdan Suditu ]